{"id":"89c5a7d3-f718-4296-bd71-267d0905b2d4","arxiv_id":"2507.22273","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"A continuum active liquid crystal model with fitted noise, diffusion, and transport parameters is shown to reproduce the organization of microtubules in human mitotic spindles.","lead":"This paper tests a coarse-grained active liquid crystal model of the mitotic spindle against electron tomography and polarization microscopy data from human cells. The model quantitatively reproduces spindle morphology and microtubule fluctuation spectra, and provides estimates of spindle material properties.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Uniform ET rescaling by psi=1.44 is load-bearing: all fitted parameters and the cross-section prediction depend on this unvalidated isotropic correction, yet post-rescaling ET and LC-PolScope geometries still differ substantially in a and b.","rationale":"The reader's weakest_assumption identifies the uniform rescaling psi = 1.44, and my analysis agrees that this is the most load-bearing assumption in the paper. The entire parameter inference and the cross-section prediction use rescaled ET coordinates, so any error in the shrinkage correction propagates into all quantitative claims. I have added a sharper piece of evidence: even after rescaling, the ET and LC-PolScope best-fit geometries disagree strongly in the long semi-axis a (8.8 vs 7.1 um) and moderately in b (5.2 vs 5.6 um). Therefore the two techniques do not simply differ by an isotropic scale factor, which weakens the uniform shrinkage justification. The concern is real but not fatal: the authors explicitly disclose the correction and provide a density cross-check (S.I. IID) whose consistency with psi = 1.44 supports the rescaling as a plausible shrinkage correction. The agreement between ET and LC-PolScope after rescaling is also checked in the correlation functions, so the claim is not obviously circular. Given that the concern is addressable with an independent shrinkage calibration or by showing the analysis is robust to alternative corrections, the reader's CONDITIONAL verdict remains appropriate. I do not see an internally inconsistent step or a mathematical error in the core derivation; the main risk is that an ad hoc global rescaling may be masking systematic differences between the two measurement modalities. The proposed test would settle whether this concern lands. Because the reader already conditioned the verdict on this assumption, my verdict is UNCHANGED.","tokens_in":1358,"tokens_out":1093,"duration_ms":132764,"concrete_test":"Recompute the full analysis pipeline (ET correlation functions, combined fits, and the x=0 cross-section Rcc) under two alternative shrinkage corrections: (i) no rescaling at all, and (ii) an anisotropic correction with independent psi_x, psi_y, psi_z calibrated from published section-compression measurements or from known dimensions of centrioles/kinetochore fibers. If the best-fit parameters D, K, v1, sc0, sn0 shift by more than their reported uncertainties, or if the ET-vs-LC-PolScope overlap in Fig. 3C and the Rcc agreement in Fig. 4C degrade, then the uniform rescaling is load-bearing and the central claim is not robust.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central quantitative claim rests on rescaling all electron tomography coordinates by a single isotropic factor psi = <d>Pol / <d>ET = 1.44 (S.I. IC). This correction forces the mean interpolar half-distance of the ET data to equal the LC-PolScope value by construction, so any comparison of overall scale between the two modalities is guaranteed in that one observable. All correlation functions, the combined fits to Eqns. (11), and the cross-section prediction Rcc(qs) in Fig. 4 are computed from rescaled coordinates. The justification for psi is freeze-substitution shrinkage, but a single scalar cannot correct non-uniform deformation, and Table S1 shows that even after rescaling the ET and LC-PolScope geometries are not reconciled: the long semi-axis differs by about 24% (8.8 vs 7.1 um) and the short semi-axis by about 7% (5.2 vs 5.6 um). If the true distortion is anisotropic or position-dependent, the rescaling will systematically distort the q-dependence of the correlation functions and the cross-sectional statistics, and the apparent quantitative agreement between ET and LC-PolScope, as well as the agreement of the model with the combined data, could be an artifact. This concern is sharpened by the fact that the same three ET spindles are used for both the long-wavelength fits and the cross-section test, so the cross-section is not an out-of-sample prediction. The concern is not that the authors are careless; the rescaling is acknowledged explicitly and a density cross-check (S.I. IID) is consistent with uniform shrinkage. But the density check is itself model-dependent, and the a/b discrepancies indicate that the two techniques measure different effective geometries even after the rescaling.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper combines serial-section electron tomography reconstructions of three HeLa metaphase spindles with LC-PolScope live imaging of eleven spindles to test an active liquid crystal model of microtubule organization. The model has deterministic equations for microtubule density and nematic director, plus independent Gaussian noise terms; the authors derive analytic spatiotemporal and equal-time correlation functions, fit five parameters to the combined ET and LC-PolScope fluctuation spectra, and then use those parameters to compute the radial density correlation in an x=0 cross-section from ET data. They also compare the predicted and measured mean orientation fields after rescaling coordinates by the interpolar half-distance. The authors report quantitative agreement between model and data for orientation, fluctuation spectra, and cross-sectional correlations at wavenumbers below about 20 rad/µm, and infer material parameters including a microtubule diffusion constant D=0.0043 µm²/s, a nematic diffusivity K=0.0021 µm²/s, and a polar transport speed v1=8.4 µm/min.","tokens_in":34360,"tokens_out":6044,"duration_ms":74134,"significance":"If the quantitative agreement holds, the paper would provide a valuable coarse-grained framework for human mitotic spindle organization, connecting local interactions, turnover, and transport to organelle-scale density and orientation fluctuations. The analytic derivation of the correlation functions in the Supporting Information is a real strength, as is the explicit combination of static ultrastructure and dynamic polarized-light data. The inference of physically interpretable parameters from spectra is a useful template for future work. However, the validation is not as strong as the text claims: the cross-section comparison uses the same three ET spindles that supplied the long-wavelength fits, and the uniform rescaling of ET coordinates is load-bearing and unvalidated. These issues do not invalidate the approach, but they require additional analysis before the predictive claims can be accepted.","major_comments":[{"comment":"The uniform rescaling of all ET coordinates by ψ=1.44 is load-bearing. Every ET correlation function in Fig. 3C and the cross-sectional Rcc(qs) in Fig. 4C are computed from rescaled coordinates, and ψ is defined by forcing the mean interpolar half-distance to match LC-PolScope. Yet Table S1 shows that the long and short semi-axes still differ between ET and LC-PolScope after rescaling (a=8.8 vs 7.1 µm, b=5.2 vs 5.6 µm), indicating that a single scalar cannot account for the systematic differences. If the underlying shrinkage is anisotropic or position-dependent, the q-dependence of the correlation functions and the fitted parameters will be distorted. The authors should provide a sensitivity analysis: repeat the fits without rescaling, with anisotropic rescaling, or with ψ varied over a plausible range, and report how the inferred parameters and the cross-section prediction change.","section":"S.I. IC, Eq. (S2); Table S1"},{"comment":"The claim that the active liquid crystal theory 'accurately predicts' the cross-sectional density fluctuations is not supported as a genuine out-of-sample prediction. The parameters (s_c^0, s_n^0, D, K, v1) are fitted to the three ET spindles from which the x=0 slice statistics are also drawn. The agreement in Fig. 4C is therefore an internal consistency check, not a prediction from independently calibrated parameters. To justify the predictive wording, the authors should use leave-one-out cross-validation on the ET spindles, or fit the parameters to LC-PolScope data alone and then test on the ET slices, and report the resulting Rcc(qs) comparison.","section":"Cross-Sections section and Fig. 4C; Materials & Methods D"},{"comment":"The LC-PolScope correlation functions are obtained from an analysis box that is displaced from the central spindle and may include a pole or the spindle boundary. The Supporting Information explicitly acknowledges that near these features the assumptions of the analytic model fail: the director is not a small perturbation about the long axis, the retardance projection formula is inaccurate, and the sample thickness is not constant. The authors argue that the signal is dominated by central-spindle microtubules, but no quantitative evidence is provided. Because the combined fits in Fig. 3C rely on these LC-PolScope spectra, the robustness of the parameter estimates to this choice of analysis box should be demonstrated, for example by recomputing the fits with a smaller, more central box or by explicitly estimating the contribution of the region where the approximations fail.","section":"S.I. III.F.1; LC-PolScope analysis box"},{"comment":"The reported value v1 = 8.4 ± 6.2 µm/min has a relative uncertainty of about 74%, so the data do not exclude v1 = 0 at a high confidence level. Since the polar transport term is central to the model's explanation of the nonzero density-director cross-correlation and of the q_y dependence of scc(q0, qy), the authors should assess and discuss the identifiability of v1 from the combined data set. A profile-likelihood or bootstrap analysis that reports confidence intervals for v1, and preferably a fit with v1 fixed to zero compared against the full model, would clarify whether polar transport is actually required by the data.","section":"Table 1 and Results: fluctuation spectra"}],"minor_comments":[{"comment":"In the expression for SNN(Q), the numerator is written as (S_C^0)^2, but based on the definition and on Main Text Eq. (11) it should be (S_N^0)^2; this is a typo in the Supporting Information.","section":"S.I. Eq. (S10)"},{"comment":"The uncertainty bands for Rcc(qs) are propagated from the ranges of the five fitted parameters, but they do not include the uncertainty in Θ, the rescaling factor ψ, or the finite number of ET spindles. The figure caption should state this limitation explicitly.","section":"Fig. 4C and Materials & Methods D"},{"comment":"The statement that data and code 'will be made available on request' is weaker than current reproducibility standards; a permanent repository or archive link would be preferable, especially since the analytic correlation functions and fitting procedures are central to the paper.","section":"Data & Code Availability"},{"comment":"The label 'LC-PolScope & EM, averaged' is ambiguous; using 'ET' consistently for electron tomography throughout the figure would avoid confusion.","section":"Figure 3C labels"}],"recommendation":"major_revision","confidential_remarks":"This is a strong and potentially influential paper, but the validation claims need to be hardened. The main issues are the unvalidated uniform rescaling of ET coordinates, the in-sample nature of the cross-section test, and the weak constraint on v1. I would not reject the paper on these grounds; all three are addressable with additional analysis within the manuscript's scope. The editor may also wish to consider whether the journal's standards require the code and data to be deposited in a permanent repository."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the real new content here is the fluctuation spectra with independent noise in density and director, plus the analytic cross-sectional correlation function. Those are worth having. The paper fits them to a combined ET and LC-PolScope dataset and extracts material parameters (D, K, v1) from human spindles in vivo. The fits look good, and the orientation-field validation against the steady-state theory is a clean replication of earlier work. Credit where due: the derivation in the SI is careful, the subsampling-based error bars are sensible, and the paper is genuinely transparent about several of its own limitations.\n\nThe soft spot is the one the stress-test flags: a single isotropic rescaling psi = 1.44 applied to all electron tomography coordinates before any correlation or cross-section analysis. That forces the mean interpolar distance to match the LC-PolScope value by construction, so it cannot serve as independent evidence of scale agreement. And even after rescaling, the long and short semi-axes still differ by roughly 24% and 7% between modalities (Table S1). If the true distortion is anisotropic or position-dependent, the rescaling will distort q-dependence and the apparent quantitative agreement could be partly artificial. The density cross-check in S.I. IID does not fully resolve this because it assumes uniform shrinkage to test uniform shrinkage. I think this is a genuine load-bearing caveat, not a manufactured one.\n\nSecond, the cross-section “prediction” in Fig. 4 uses the same three ET spindles from which the long-wavelength parameters were fitted. It is a consistency check, not an out-of-sample test. The paper says “without additional fitting parameters,” which is technically true, but the word “prediction” overstates it. Third, no code or data are deposited; “available on request” is a real inconvenience for a paper with this much analysis.\n\nNone of this kills the central claim. The model does plausibly explain the fluctuation spectra, and the parameter inference is a useful step. But the quantitative strength of that claim is lower than the abstract implies. The paper deserves a serious referee and would benefit from independent validation data or a more careful treatment of the rescaling. I would send it to peer review with the expectation of major revisions focused on the cross-section test and the rescaling.\n\nFor a reading group, it is a good discussion piece on the gap between continuum theory and spindle data. I would not cite it for the cross-section prediction, but I might cite it for the correlation function derivation and parameter values.","headline":"A serious, mostly sound extension of active liquid crystal theory to human spindle fluctuations, but the uniform ET rescaling and same-spindle cross-section test make the quantitative claim less secure than it looks.","tokens_in":34927,"tokens_out":1607,"would_cite":true,"duration_ms":22928,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A minimal active liquid crystal model, with parameters inferred from combined electron tomography and live polarized-light microscopy, quantitatively reproduces both the fluctuation spectra and the metaphase-plate cross-sectional packing…","keywords":["active liquid crystal","mitotic spindle","microtubule self-organization","electron tomography","LC-PolScope","fluctuation spectra","nematic elasticity","HeLa cells"],"falsifier":"Perform electron tomography on a spindle containing embedded fiduciary markers of known spacing, or image the same spindle live with LC-PolScope and then after fixation and embedding, to measure shrinkage locally rather than by a single global factor; if a uniform rescaling cannot bring all coordinates into agreement with the live geometry, the fitted parameters and the parameter-free cross-sectional prediction would not survive the corrected coordinates.","tokens_in":2148,"feed_emoji":"🧬","tokens_out":1995,"duration_ms":81278,"temperature":0.7,"pith_summary":"This paper argues that the collective organization of thousands of microtubules in a human mitotic spindle can be captured by a minimal continuum theory borrowed from active liquid crystal physics. By combining static, nanometer-resolution electron tomography reconstructions with dynamic, optical-resolution LC-PolScope measurements of the same cell type, the authors show that a handful of physically interpretable parameters (a diffusion constant, a nematic elasticity, a polar transport speed, and two noise amplitudes) reproduce the measured density and orientation fluctuation spectra. More strongly, the same parameters, without further fitting, predict the arrangement of microtubules in a spindle cross-section at the metaphase plate down to length scales of roughly 300 nanometers. If correct, this means that spindle self-organization does not require specialized bundle-forming mechanisms on those scales; apparent bundles are transient density fluctuations of the active nematic, and the spindle's material properties can be measured in living cells.","feed_headline":"Liquid crystal theory predicts spindle microtubule packing","feed_subtitle":"Electron tomography plus polarized light show a few parameters capture spindle fluctuations and cross-sections down to 0.3 µm.","key_machinery":"The central object is a minimal active nematic field theory: a pair of stochastic partial differential equations for the microtubule density $\\rho(R,t)$ and the nematic director $\\hat{N}(R,t)$, Eqns. (3) and (4). The density equation has turnover ($\\Gamma_0$ nucleation, $\\Theta$ catastrophe rate), diffusive-like motion with diffusivity $D$, and a symmetry-breaking polar transport term $v_1 \\rho \\hat{N}$; the director equation is a projected diffusion equation with nematic diffusivity $K$. The theory is validated by linearizing about a uniform, aligned steady state and computing the Fourier-space correlation functions of density and orientation fluctuations (Eqns. (10) and (11)), including independent white noises in both fields. The cross-sectional prediction at the metaphase plate comes from marginalizing the 3D density structure factor over the spindle-axis wavevector, yielding the closed-form expression $R_{CC}(q_s)$ in Eqn. (13), which is then compared with electron tomography-derived slice data without any new fitting parameters.","core_discovery":"The central discovery is that the spatiotemporal statistics of microtubule density and orientation in human mitotic spindles are quantitatively described by a linearized active nematic model with two coupled fields: a density field obeying a diffusion-advection equation with turnover and a polar transport term $v_1 \\hat{N}$, and a director field relaxing by nematic elasticity $K$. Including independent Gaussian noise sources in both fields, the model's Fourier-space correlation functions (Eqns. (10) and (11)) fit the combined electron tomography and LC-PolScope data with parameters that are physically interpretable, such as $D = (0.0043 \\pm 0.0023)\\,\\mu\\text{m}^2/\\text{s}$ and $K = (0.0021 \\pm 0.0002)\\,\\mu\\text{m}^2/\\text{s}$. The paper then shows that the same fitted parameters, without additional adjustment, predict the radial density correlation function $R_{CC}(q_s)$ of microtubule intersections in a constant-$x$ cross-section near the metaphase plate, accurate for wavenumbers up to $q_s^* \\approx 20\\,\\text{rad}\\,\\mu\\text{m}^{-1}$ (real-space distances greater than about $0.3\\,\\mu\\text{m}$). The authors interpret this as evidence that local interactions, diffusive-like motion, and polar transport govern the spindle's microtubule network, and that observed microtubule bundles at these scales are transient density fluctuations rather than structures requiring dedicated bundling machinery.","pith_inferences":["A direct test of the paper's interpretation would be to track individual microtubule bundles over time in living spindles: if bundles are transient density fluctuations, their lifetimes and spatial scales should match the relaxation rates predicted by the fitted diffusivity and nematic elasticity, whereas stable bundles would imply missing physics.","The necessity of adding independent noise to the density equation, a modification relative to earlier models that only included orientation noise, suggests that density fluctuations are not slaved to orientation noise; this could be probed by pharmacologically altering microtubule nucleation or turnover and checking whether the density noise amplitude changes independently of the orientation noise","If the theory is right, the cross-sectional prediction $R_{CC}(q_s)$ should break down reproducibly at wavenumbers beyond $q_s^*$ where molecular cross-linker spacing becomes relevant; comparing the precise location of that breakdown across cell lines or after cross-linker perturbations could offer a quantitative readout of molecular-scale organization.","The uniform shrinkage rescaling used to bring electron tomography and LC-PolScope geometries into agreement is the least controlled step; validating it with independent fiduciary markers would determine whether the claimed parameter-free cross-sectional prediction holds beyond the heuristic correction."],"forward_implications":["Spindle material properties such as microtubule diffusivity, nematic elasticity, polar transport speed, and fluctuation noise amplitudes can be measured in vivo from correlation functions, providing a quantitative link between molecular interactions and mesoscale spindle behavior.","Apparent microtubule bundles in the metaphase plate, at length scales above roughly 0.3 micrometers, are explained as transient density fluctuations of the active nematic, without invoking dedicated bundle-forming factors.","The measured parameters imply that along the spindle long axis, density fluctuations relax mostly by diffusive-like motion decoupled from orientation, while along the short axis, density dynamics are coupled to the director through the active polar transport term, consistent with the observed nonzero director-density cross-correlation.","The same coarse-grained framework can be applied to other cell types and experimental perturbations, allowing systematic comparison of spindle physics across conditions.","The success of this parameter-free cross-sectional prediction suggests that the theory can be used to infer material properties of the spindle from a single static electron tomography reconstruction, as long as the long-wavelength fluctuation parameters are known."],"supporting_citations":[{"why":"Provides the original active liquid crystal model of spindle self-organization and the earlier analysis of fluctuations in Xenopus extract spindles that this paper extends by adding independent density noise.","marker":"[10]"},{"why":"Establishes that the steady-state director equation with spindle-geometry anchoring describes HeLa spindle orientation, and supplies the microtubule turnover rate Theta = 0.047 per second used as a fixed parameter.","marker":"[13]"},{"why":"Supplies the three complete electron tomography reconstructions of HeLa spindles used for all static microtubule position and orientation analyses.","marker":"[16]"},{"why":"Provides the per-microtubule retardance constant A0 and the calibration approach used to estimate the LC-PolScope optical impulse function for deconvolving correlation functions.","marker":"[22]"},{"why":"Forms the theoretical basis of active liquid crystal hydrodynamics from which the coarse-grained equations of motion are derived.","marker":"[18]"},{"why":"Supplies the standard statistical mechanics formalism (Fourier-space correlation functions) used to estimate and compare density and orientation fluctuations.","marker":"[6]"},{"why":"Documents section collapse and shrinkage in electron microscopy sample preparation, motivating the uniform rescaling factor psi applied to electron tomography coordinates.","marker":"[42]"}],"fun_headline_variants":["Active nematic model explains spindle microtubule order","Spindle microtubules obey liquid crystal dynamics","One model fits spindle images and fluctuations","Microtubule patterns in spindles predicted by theory","Liquid crystal theory casts spindle organization precisely"],"cache_read_input_tokens":36864,"weakest_assumption_plain":"The quantitative agreement between electron tomography and polarized-light microscopy rests on a single scalar rescaling of all electron-tomography coordinates (shrinkage factor approximately 1.44) chosen to force the mean pole spacing to match the LC-PolScope value; if shrinkage or geometric distortion is non-uniform, or if the two techniques measure systematically different spindle geometries, the apparent agreement and the fitted parameters could be artifacts.","fun_headline_variants_meta":{"raw":{"variants":["Active nematic model explains spindle microtubule order","Spindle microtubules obey liquid crystal dynamics","One model fits spindle images and fluctuations","Microtubule patterns in spindles predicted by theory","Liquid crystal theory casts spindle organization precisely"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000256,"raw_usage":{"total_tokens":1597,"prompt_tokens":988,"completion_tokens":609,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":604,"completion_tokens_details":{"reasoning_tokens":540}},"tokens_in":604,"tokens_out":609,"duration_ms":8355,"temperature":1.0,"reasoning_tokens":540,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T11:52:07.092372+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Perform electron tomography on a spindle containing embedded fiduciary markers of known spacing, or image the same spindle live with LC-PolScope and then after fixation and embedding, to measure shrinkage locally rather than by a single global factor; if a uniform rescaling cannot bring all coordinates into agreement with the live geometry, the fitted parameters and the parameter-free cross-sectional prediction would not survive the corrected coordinates.","supporting_citations":[{"cited_title":"Instabilities and pattern formation in active particle suspensions: Kinetic theory and continuum simulations.Physical Review Letters, 100(17):178103, 2008","cited_arxiv_id":null,"evidence_quote":"Provides the original active liquid crystal model of spindle self-organization and the earlier analysis of fluctuations in Xenopus extract spindles that this paper extends by adding independent density noise."},{"cited_title":"Long-range repulsion between chromosomes in mammalian oocyte spindles","cited_arxiv_id":null,"evidence_quote":"Establishes that the steady-state director equation with spindle-geometry anchoring describes HeLa spindle orientation, and supplies the microtubule turnover rate Theta = 0.047 per second used as a fixed parameter."},{"cited_title":"Nucleoporin levels regulate cell cycle progression and phase-specific gene expression","cited_arxiv_id":null,"evidence_quote":"Supplies the three complete electron tomography reconstructions of HeLa spindles used for all static microtubule position and orientation analyses."},{"cited_title":"A reliable, noninvasive technique for spindle imaging and enucleation of mammalian oocytes.Nature biotechnology, 18(2):223–225, 2000","cited_arxiv_id":null,"evidence_quote":"Provides the per-microtubule retardance constant A0 and the calibration approach used to estimate the LC-PolScope optical impulse function for deconvolving correlation functions."},{"cited_title":"Using 3d large scale tomography to study force generation in the mitotic spindle, 2024","cited_arxiv_id":null,"evidence_quote":"Forms the theoretical basis of active liquid crystal hydrodynamics from which the coarse-grained equations of motion are derived."},{"cited_title":"Deep proteomics of the xenopus laevis egg using an mrna-derived reference database.Current biology, 24(13):1467–1475, 2014","cited_arxiv_id":null,"evidence_quote":"Supplies the standard statistical mechanics formalism (Fourier-space correlation functions) used to estimate and compare density and orientation fluctuations."},{"cited_title":"Resampling fewer than n observations: gains, losses, and remedies for losses","cited_arxiv_id":null,"evidence_quote":"Documents section collapse and shrinkage in electron microscopy sample preparation, motivating the uniform rescaling factor psi applied to electron tomography coordinates."}],"review_version":1}